Chuong 4 3 LAP BAN DO HIEN TRANG SDD

BÀI 3: LẬP BẢN ĐỒ LỚP PHỦ/BĐ SỬ DỤNG ĐẤT 3.1 Hiển thị ảnh - DISPLAYING IMAGES 3.2 Tạo ảnh tổ hợp màu - CRREATING A COMPOSITE IMAGE 3.3 Tăng cường độ tương phản ảnh - IMAGE ENHANCEMENT 3.4 Phân lớp không kiểm định - UNSUPERVISED CLASSIFICATION 3.5 Phân lớp có kiểm định - SUPERVISED CLASSIFICATION 3.6 Đánh giá độ xác - ACCURACY ASSESSMENT DUONG DANG KHOI 3.1 HIỂN THỊ ẢNH - Ảnh Landsat thu hiển thị thang màu xám (grayscale), mở arcGIS liệt kê Table of Contents ArcMap - Mỗi kênh hữu ích cho nhận biết số đối tượng: DUONG DANG KHOI 3.1.1 Mở tệp ảnh ArcMap Khi giải nén thư mục tải về, ta có tệp ảnh sau: DUONG DANG KHOI Mở ArcMap mở tệp trắng Chọn toàn tệp ảnh giải nén Mở tất file chọn sau: DUONG DANG KHOI DUONG DANG KHOI DUONG DANG KHOI Kiểm tra thông tin file ảnh DUONG DANG KHOI DUONG DANG KHOI 3.2 TẠO ẢNH TỔ HỢP (CREATING A COMPOSITE IMAGE) - Ảnh Landsat thu thập lưu dạng thang màu xám (grayscale) kênh lưu trữ riêng - Bằng việc tạo ảnh tổ hợp, ta hiển thị ảnh dạng màu, hiển thị ảnh theo định dạng màu RGB chp phep ta quan sát thấy đối tượng khác ảnh, đặc biệt nhóm đối tượng lớp phủ: urban areas, forests, agriculture, water bodies DUONG DANG KHOI - Mở ArcMap, mở a new map document, mở toàn tệp ảnh Landsat vào new map document - Thông qua kết hợp ảnh, ta hiển thị kênh khác sử dụng khung màu RGB DUONG DANG KHOI 10 Identifying Each Random Point’s Value from Aerial Imagery - We are going to use Google Earth as the reference image with which to identify each random point’s value - To this, we need to export the file into kml file so we can view it in Google Earth - Go to your Toolbox, Tools/To KML/Layer to KML DUONG DANG KHOI 91 - Enter your random points file as the Layer and then name your new kml file DUONG DANG KHOI 92 - Open Google Earth Once it is open, go to File/Open and navigate to your workspace Double-click on your kml file and then click open to add it to Google Earth DUONG DANG KHOI 93 Adding the reference value to your Random Points shapefile - You will need to add a field in your Attribute Table in ArcMap so you can enter each point’s value - After identifying values for each point by examining the aerial photos, you will have a reference set against which to compare your classifications DUONG DANG KHOI 94 - The Attribute Table for your random points file and the reference image should look like this DUONG DANG KHOI 95 - Now we need to add the values for the classified images Input Points Feature is your random points file Input Raster is your Classified Image DUONG DANG KHOI 96 - Open the Attribute Table of your new file ArcGIS has added a new field and called it RASTERVALU - It pulled the values of the cells that match to each random point and added them to the Attribute Table - We have added a new field and called it ISO so we can transfer the values from RASTERVALU DUONG DANG KHOI 97 - Left-click on the ISO column and select Field Calculator To transfer the values from RASTERVALU, double-click on RASTERVALU in the Fields box of the screen, which populates it into the equation below Click OK DUONG DANG KHOI 98 DUONG DANG KHOI 99 DUONG DANG KHOI 100 Compiling the Error matrix Table: Error matrix DUONG DANG KHOI 101 Calculating Overall Accuracy - Overall accuracy is the percentage of random points that are the same in both images For our matrix above, that is 18 points (1 water, urban, 14 forest and agriculture) We have a total of 25 random points, so our overall accuracy is 18/25 or 72% - User’s accuracy, Producer’s accuracy, errors of omission and errors of commission can also be calculated from the error matrix DUONG DANG KHOI 102 Calculation of Cohen’s Kappa Kappa provides us with insight into our classification scheme and whether or not we achieved results better than we would have achieved strictly by chance The formula for kappa is: - Observed is overall accuracy - Expected is calculated from the rows and column totals DUONG DANG KHOI 103 First, you calculate the product of the rows and columns Total row Total column DUONG DANG KHOI 104 DUONG DANG KHOI 105 ... coniferous forest being darker green than deciduous DUONG DANG KHOI 13 DUONG DANG KHOI 14 DUONG DANG KHOI 15 3. 3 TĂNG CƯỜNG ẢNH 3. 3.1 Radiometric Enhancement Using ArcGIS as a Viewer for Radiometric... màu RGB DUONG DANG KHOI 10 DUONG DANG KHOI 11 RGB: 4- 3- 2 HÀ NỘI The standard "false color" composite Vegetation appears in shades of red, urban areas are cyan blue, and soils vary from dark to... Densely populated urban areas are shown in light blue This TM band combination gives results similar to traditional color infrared aerial photography DUONG DANG KHOI 12 RGB: 7 -4- 2 This combination